You will get a fixed PyTorch training run (CUDA aligned) in 24–48h
Rising Talent
Rising Talent
Project details
I fix broken PyTorch training runs fast (usually within 24–48h). I’ll reproduce your error, align the Torch ↔ CUDA ↔ driver matrix, create a clean venv/conda environment with pinned versions, and deliver a verified training command plus clear notes to keep it stable.
You’ll receive: requirements.txt/environment.yml, a short log of the successful run, and a README with "stability notes" and a checklist. In the Standard/Premium tiers I also include tidy environment setup and, for Premium, a one-command CI script and optional W&B/CSV metrics logging.
This service targets environment/dependency/runtime issues on NVIDIA GPUs (Ubuntu or Windows). I work chat-only (no calls), respect your privacy, and can sign an NDA if needed.
You’ll receive: requirements.txt/environment.yml, a short log of the successful run, and a README with "stability notes" and a checklist. In the Standard/Premium tiers I also include tidy environment setup and, for Premium, a one-command CI script and optional W&B/CSV metrics logging.
This service targets environment/dependency/runtime issues on NVIDIA GPUs (Ubuntu or Windows). I work chat-only (no calls), respect your privacy, and can sign an NDA if needed.
Machine Learning Tools
Keras, MLflow, NumPy, NVIDIA AI Platform, OpenCV, pandas, Python, PyTorch, scikit-learn, SciPy, TensorFlowWhat's included
| Service Tiers |
Starter
$99
|
Standard
$199
|
Advanced
$349
|
|---|---|---|---|
| Delivery Time | 1 day | 2 days | 3 days |
Number of Revisions | 1 | 2 | 3 |
Number of Model Variations | 0 | 0 | 0 |
Number of Scenarios | 0 | 0 | 0 |
Number of Graphs/Charts | 0 | 0 | 1 |
Model Validation/Testing | |||
Model Documentation | - | ||
Data Source Connectivity | - | - | - |
Source Code | - |
Optional add-ons
You can add these on the next page.
Additional Revision
+$49
Additional Graph/Chart
+$25
Source Code
+$99
12h Rush Delivery (custom add-on)
+$79
Extra Validation/Eval run
+$79Frequently asked questions
About Dejan
PhD, Senior ML Engineer | YOLO+SAHI CV | LLMs (RAG, QLoRA)
Vranje, Serbia - 11:00 pm local time
• Computer Vision: YOLOv5/8 + SAHI + WBF for small-object UAV imagery; +24% mAP@0.50:0.95 with FN↓ at a defined operating point. Real-time PyTorch/OpenCV/CUDA pipelines.
• NLP / LLMs: Led a bilingual GPT-2–class model effort; QLoRA fine-tuning, RAG (Elasticsearch/FAISS), cross-encoder reranking, evaluation (BLEU/ROUGE/BERTScore).
• Ops & Quality: Reproducible repos (Docker, W&B, CI), deterministic runs, clear metric reports. Own GPU server.
• Domains: Legal-tech, safety-critical CV (UAV/remote sensing), time-series forecasting, academia.
• Deliverables: one-command run + configs, README with stability notes, PR curves & tables, short metric brief, post-delivery support window.
• Collaboration: Upwork Messages (chat-only), fixed-price milestones for speed & predictability. Availability: 30+ hrs/week, 24–48h turnaround per revision wave.
Steps for completing your project
After purchasing the project, send requirements so Dejan can start the project.
Delivery time starts when Dejan receives requirements from you.
Dejan works on your project following the steps below.
Revisions may occur after the delivery date.
Intake & baseline reproduce
I receive the repo/zip, environment details and error log. I run a baseline to reproduce the issue and confirm a minimal repro.
Diagnosis (Torch/CUDA/driver matrix)
I check Torch↔CUDA↔driver compatibility, lock the versions, and share a short fix plan in chat.